I think it was Jonathan Haidt who said something like this: Individuals are irrational, but systems can be designed to produce rational outcomes.

Science is the best example. Most scientist can be very irrational in many ways, but when they are forced to act in a scientific environment, the best arguments win in the long run.